In [3]:
# Read Image
image_color = imread("Sample03/sofa.jpg")
# Convert Image into Gray
image_gray = cv2.cvtColor(image_color, cv2.COLOR_RGB2GRAY)
# Display Image
ShowTwoImages(image_color, image_gray)
import numpy as np
import pandas as pd
import cv2
from matplotlib import pyplot as plt
from pylab import imread, imsave
from skimage.color import rgb2gray
def imshows(ImageData, LabelData, rows, cols, gridType = False):
# Convert ImageData and LabelData to List
from matplotlib import pyplot as plt
ImageArray = list(ImageData)
LabelArray = list(LabelData)
if(rows == 1 & cols == 1):
fig = plt.figure(figsize=(20,20))
else:
fig = plt.figure(figsize=(cols*8,rows*5))
for i in range(1, cols * rows + 1):
fig.add_subplot(rows, cols, i)
image = ImageArray[i - 1]
# If the channel number is less than 3, we display as grayscale image
# otherwise, we display as color image
if (len(image.shape) < 3):
plt.imshow(image, plt.cm.gray)
plt.grid(gridType)
else:
plt.imshow(image)
plt.grid(gridType)
plt.title(LabelArray[i - 1])
plt.show()
def ShowThreeImages(IM1, IM2, IM3):
imshows([IM1, IM2, IM3], ["Image 1","Image 2", "Image 3"], 1, 3)
def ShowTwoImages(IM1, IM2):
imshows([IM1, IM2], ["Image 1","Image 2"], 1, 2)
def ShowOneImage(IM):
imshows([IM], ["Image"], 1, 1)
def ShowListImages(listImage, row, col):
listCaption = []
for i in range(len(listImage)):
listCaption.append(str(i))
imshows(listImage,listCaption,row,col)
# Read Image
image_color = imread("Sample03/sofa.jpg")
# Convert Image into Gray
image_gray = cv2.cvtColor(image_color, cv2.COLOR_RGB2GRAY)
# Display Image
ShowTwoImages(image_color, image_gray)
kernel_3_3 = np.ones((3,3),np.float32)/9
kernel_5_5 = np.ones((5,5),np.float32)/25
print("Filtering Kernel 3x3:")
display(kernel_3_3)
print("Filtering Kernel 5x5:")
display(kernel_5_5)
image_filter_3_3_01 = cv2.filter2D(image_color,-1,kernel_3_3)
image_filter_3_3_02 = cv2.filter2D(image_filter_3_3_01,-1,kernel_3_3)
image_filter_5_5_01 = cv2.filter2D(image_color,-1,kernel_5_5)
image_filter_5_5_02 = cv2.filter2D(image_filter_5_5_01,-1,kernel_5_5)
ShowThreeImages(image_color, image_filter_3_3_01, image_filter_3_3_02)
ShowThreeImages(image_color, image_filter_5_5_01, image_filter_5_5_02)
Filtering Kernel 3x3:
array([[0.11111111, 0.11111111, 0.11111111],
[0.11111111, 0.11111111, 0.11111111],
[0.11111111, 0.11111111, 0.11111111]], dtype=float32)
Filtering Kernel 5x5:
array([[0.04, 0.04, 0.04, 0.04, 0.04],
[0.04, 0.04, 0.04, 0.04, 0.04],
[0.04, 0.04, 0.04, 0.04, 0.04],
[0.04, 0.04, 0.04, 0.04, 0.04],
[0.04, 0.04, 0.04, 0.04, 0.04]], dtype=float32)
def variance_of_laplacian(image):
# compute the Laplacian of the image and then return the focus
# measure, which is simply the variance of the Laplacian
return cv2.Laplacian(image, cv2.CV_64F).var()
blur_mesurement = variance_of_laplacian(image_color)
blur_mesurement_3_3_01 = variance_of_laplacian(image_filter_3_3_01)
blur_mesurement_3_3_02 = variance_of_laplacian(image_filter_3_3_02)
print("Blur Measurement of image_color:", blur_mesurement)
print("Blur Measurement of image_filter_3_3_01:", blur_mesurement_3_3_01)
print("Blur Measurement of image_filter_3_3_02:", blur_mesurement_3_3_02)
Blur Measurement of image_color: 424.2608431387909 Blur Measurement of image_filter_3_3_01: 32.49074150718744 Blur Measurement of image_filter_3_3_02: 12.51062115894779
blur_mesurement = variance_of_laplacian(image_color)
blur_mesurement_5_5_01 = variance_of_laplacian(image_filter_5_5_01)
blur_mesurement_5_5_02 = variance_of_laplacian(image_filter_5_5_02)
print("Blur Measurement of image_color:", blur_mesurement)
print("Blur Measurement of image_filter_5_5_01:", blur_mesurement_5_5_01)
print("Blur Measurement of image_filter_5_5_02:", blur_mesurement_5_5_02)
Blur Measurement of image_color: 424.2608431387909 Blur Measurement of image_filter_5_5_01: 10.57958835677548 Blur Measurement of image_filter_5_5_02: 3.415116354886171
text = "Blurry measurement"
fm = blur_mesurement_5_5_02
image_filter_5_5_02_text = image_filter_5_5_02.copy()
cv2.putText(image_filter_5_5_02_text, "{}: {:.2f}".format(text, fm), (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.8,
(255, 0, 0), 2)
ShowTwoImages(image_filter_5_5_02, image_filter_5_5_02_text)
# Read Image
image_color = imread("Sample03/flower_2.jpg")
# Convert Image into Gray
image_gray = cv2.cvtColor(image_color, cv2.COLOR_RGB2GRAY)
# Display Image
ShowTwoImages(image_color, image_gray)
kernel_sharpen_01 = np.array([[-1,-1,-1],
[-1, 9,-1],
[-1,-1,-1]])
kernel_sharpen_02 =np.array([[-1,-1,-1,-1,-1],
[-1,-1,-1,-1,-1],
[-1, -1,25,-1,-1],
[-1,-1,-1,-1,-1],
[-1,-1,-1,-1,-1]])
print("Filtering kernel_sharpen_01:")
display(kernel_sharpen_01)
print("Filtering kernel_sharpen_02:")
display(kernel_sharpen_02)
image_color_sharpen_01 = cv2.filter2D(image_color, -1, kernel_sharpen_01)
image_gray_sharpen_01 = cv2.filter2D(image_gray, -1, kernel_sharpen_01)
image_color_sharpen_02 = cv2.filter2D(image_color, -1, kernel_sharpen_02)
image_gray_sharpen_02 = cv2.filter2D(image_gray, -1, kernel_sharpen_02)
ShowThreeImages(image_color, image_color_sharpen_01, image_color_sharpen_02)
ShowThreeImages(image_gray, image_gray_sharpen_01, image_gray_sharpen_02)
Filtering kernel_sharpen_01:
array([[-1, -1, -1],
[-1, 9, -1],
[-1, -1, -1]])
Filtering kernel_sharpen_02:
array([[-1, -1, -1, -1, -1],
[-1, -1, -1, -1, -1],
[-1, -1, 25, -1, -1],
[-1, -1, -1, -1, -1],
[-1, -1, -1, -1, -1]])
print("Sharpen Measurement of image_color:", variance_of_laplacian(image_color))
print("Sharpen Measurement of image_color_sharpen_01:", variance_of_laplacian(image_color_sharpen_01))
print("Sharpen Measurement of image_color_sharpen_01:", variance_of_laplacian(image_color_sharpen_02))
print("Sharpen Measurement of image_gray:", variance_of_laplacian(image_gray))
print("Sharpen Measurement of image_gray_sharpen_01:", variance_of_laplacian(image_gray_sharpen_01))
print("Sharpen Measurement of image_gray_sharpen_01:", variance_of_laplacian(image_gray_sharpen_02))
Sharpen Measurement of image_color: 458.4116327940773 Sharpen Measurement of image_color_sharpen_01: 11170.959449601622 Sharpen Measurement of image_color_sharpen_01: 29411.538459131363 Sharpen Measurement of image_gray: 367.6566636404424 Sharpen Measurement of image_gray_sharpen_01: 10938.927339722139 Sharpen Measurement of image_gray_sharpen_01: 25540.133050838063
text = "Sharpen Measurement"
fm = variance_of_laplacian(image_color_sharpen_02)
image_color_sharpen_02_text = image_color_sharpen_02.copy()
cv2.putText(image_color_sharpen_02_text, "{}: {:.2f}".format(text, fm), (10, 50), cv2.FONT_HERSHEY_SIMPLEX, 0.8,
(255, 0, 0), 2)
ShowTwoImages(image_color_sharpen_02, image_color_sharpen_02_text)
import cv2
import os
def load_images_from_folder(path):
Join = os.path.join
files = [Join(path,f) for f in os.listdir(path)]
images = []
for path in files:
images.append(imread(path))
return images
path="Image_Input"
list_images = load_images_from_folder(path)
ShowListImages(list_images,len(list_images)//3,3)
for i in range(len(list_images)):
print("Blur Measurement of image_color {}: {}".format(i, variance_of_laplacian(list_images[i])))
Blur Measurement of image_color 0: 460.12285089856954 Blur Measurement of image_color 1: 377.37529522721206 Blur Measurement of image_color 2: 2272.323747420017 Blur Measurement of image_color 3: 2516.730283159722 Blur Measurement of image_color 4: 587.4574180909062 Blur Measurement of image_color 5: 325.34208639219986 Blur Measurement of image_color 6: 424.2608431387909
blur_threshold = 500
text_images = []
for i in range(len(list_images)):
if variance_of_laplacian(list_images[i]) > blur_threshold:
img_text = list_images[i].copy()
cv2.putText(img_text, "{}".format("Good Image"), (10, 30), \
cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255, 0, 0), 2)
text_images.append(img_text)
else:
img_text = list_images[i].copy()
cv2.putText(img_text, "{}: {:.2f}".format("Blur Image", variance_of_laplacian(list_images[i])), (10, 30), \
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (255, 0, 0), 2)
text_images.append(img_text)
ShowListImages(text_images, 2, 3)
path = 'Image_BlurDetection_Output'
os.mkdir(path)
for i in range(len(text_images)):
imsave(os.path.join(path, '{}.jpg'.format(i)), text_images[i])
# Read Image
image_color = imread("Sample03/house_2.jpg")
# Convert Image into Gray
image_gray = cv2.cvtColor(image_color, cv2.COLOR_RGB2GRAY)
# Display Image
ShowTwoImages(image_color, image_gray)
def max_rgb_filter(image):
# split the image into its BGR components
(B, G, R) = cv2.split(image)
# find the maximum pixel intensity values for each
# (x, y)-coordinate,, then set all pixel values less
# than M to zero
M = np.maximum(np.maximum(R, G), B)
R[R < M] = 0
G[G < M] = 0
B[B < M] = 0
# merge the channels back together and return the image
return cv2.merge([B, G, R])
image_color_rgbmax = max_rgb_filter(image_color)
ShowTwoImages(image_color, image_color_rgbmax)
def SegmentColorImageByMask(IM, Mask):
Mask = Mask.astype(np.uint8)
result = cv2.bitwise_and(IM, IM, mask = Mask)
return result
image_maxR_mask = image_gray < 0
image_maxG_mask = image_gray < 0
image_maxB_mask = image_gray < 0
R = image_color_rgbmax[:,:,0]
G = image_color_rgbmax[:,:,1]
B = image_color_rgbmax[:,:,2]
image_maxR_mask[(G == 0) & (B == 0)] = 1
image_maxG_mask[(R == 0) & (B == 0)] = 1
image_maxB_mask[(G == 0) & (R == 0)] = 1
image_maxR = SegmentColorImageByMask(image_color, image_maxR_mask)
image_maxG = SegmentColorImageByMask(image_color, image_maxG_mask)
image_maxB = SegmentColorImageByMask(image_color, image_maxB_mask)
ShowTwoImages(image_color, image_color_rgbmax)
ShowThreeImages(image_maxR_mask, image_maxG_mask, image_maxB_mask)
ShowThreeImages(image_maxR, image_maxG, image_maxB)
# Use function OpenCV
print("Blur Measurement of image_color:", variance_of_laplacian(image_color))
image_color_blur_02 = cv2.GaussianBlur(image_color,(5,5),0)
image_color_sharpen_02 = cv2.filter2D(image_color_blur_02, -1, kernel_sharpen_02)
print("Blur Measurement of image_color after blur and sharpen:", variance_of_laplacian(image_color_sharpen_02))
Blur Measurement of image_color: 2516.730283159722 Blur Measurement of image_color after blur and sharpen: 16298.072266227779
print("Blur Measurement of image_color:", variance_of_laplacian(image_color))
image_color_sharpen_02 = cv2.filter2D(image_color, -1, kernel_sharpen_02)
image_color_blur_02 = cv2.GaussianBlur(image_color_sharpen_02,(5,5),0)
print("Blur Measurement of image_color after sharpen and blur:", variance_of_laplacian(image_color_blur_02))
Blur Measurement of image_color: 2516.730283159722 Blur Measurement of image_color after sharpen and blur: 1440.0133764517575
# construct the Laplacian kernel used to detect edge-like
# regions of an image
# construct the Sobel x-axis kernel
kernel_sobelX = np.array((
[-1, 0, 1],
[-2, 0, 2],
[-1, 0, 1]), dtype="int")
# construct the Sobel y-axis kernel
kernel_sobelY = np.array((
[-1, -2, -1],
[0, 0, 0],
[1, 2, 1]), dtype="int")
image_color_edge_sobelX = cv2.filter2D(image_color, -1, kernel_sobelX)
image_color_edge_sobelY = cv2.filter2D(image_color, -1, kernel_sobelY)
image_color_edge = image_color_edge_sobelX + image_color_edge_sobelY
ShowTwoImages(image_color, image_color_edge)
ShowTwoImages(image_color_edge_sobelX, image_color_edge_sobelY)
image_gray_edge_sobelX = cv2.filter2D(image_gray, -1, kernel_sobelX)
image_gray_edge_sobelY = cv2.filter2D(image_gray, -1, kernel_sobelY)
image_gray_edge = image_gray_edge_sobelX + image_gray_edge_sobelY
ShowTwoImages(image_gray, image_gray_edge)
ShowTwoImages(image_gray_edge_sobelX, image_gray_edge_sobelY)
